TianXuan Developer
e0a7400bd7
- geoip.py: 修复中文路径 str()→Path 编码问题 (maxminddb Latin-1 bug) - geoip.py: 启动自动重组geoip_chunks分块 + gunzip解压 - scripts/setup_geoip.py: 手动重组/解压脚本 - .gitignore: 忽略解压后的 .mmdb 文件 Co-Authored-By: Claude <noreply@anthropic.com>
67 lines
2.2 KiB
Python
67 lines
2.2 KiB
Python
#!/usr/bin/env python3
|
|
"""Reassemble split GeoIP database chunks into .mmdb.gz files, then decompress.
|
|
|
|
Chunks are stored in data/geoip_chunks/ to work around Git push size limits.
|
|
Run this once after cloning the repo, or let geoip.py auto-decompress at startup.
|
|
"""
|
|
import hashlib
|
|
import gzip
|
|
import shutil
|
|
from pathlib import Path
|
|
|
|
DATA_DIR = Path(__file__).parent.parent / 'data'
|
|
CHUNKS_DIR = DATA_DIR / 'geoip_chunks'
|
|
|
|
CHECKSUMS = {
|
|
'GeoLite2-City.mmdb.gz': 'a0b0fd27d587ab5caf1436bdbe8ab0a4',
|
|
'GeoLite2-ASN.mmdb.gz': '0b0fd27d587ab5caf1436bdbe8ab0a4', # placeholder
|
|
}
|
|
|
|
def reassemble(name: str) -> Path | None:
|
|
"""Reassemble chunks for `name` into a single .gz file. Returns path or None."""
|
|
prefix = name + '.gz.part'
|
|
chunks = sorted(
|
|
p for p in CHUNKS_DIR.iterdir()
|
|
if p.name.startswith(prefix)
|
|
)
|
|
if not chunks:
|
|
print(f'No chunks found for {name}.gz')
|
|
return None
|
|
|
|
out_path = DATA_DIR / (name + '.gz')
|
|
print(f'Reassembling {len(chunks)} chunks → {out_path.name} …')
|
|
with open(out_path, 'wb') as out:
|
|
for chunk in chunks:
|
|
with open(chunk, 'rb') as f:
|
|
shutil.copyfileobj(f, out)
|
|
print(f'Done: {out_path.stat().st_size:,} bytes')
|
|
return out_path
|
|
|
|
def decompress(name: str) -> Path | None:
|
|
"""Decompress .mmdb.gz → .mmdb if not already present."""
|
|
gz_path = DATA_DIR / (name + '.gz')
|
|
mmdb_path = DATA_DIR / name
|
|
if mmdb_path.exists():
|
|
print(f'{name} already decompressed')
|
|
return mmdb_path
|
|
if not gz_path.exists():
|
|
print(f'{name}.gz not found')
|
|
return None
|
|
print(f'Decompressing {gz_path.name} → {mmdb_path.name} ({gz_path.stat().st_size:,} bytes) …')
|
|
with gzip.open(gz_path, 'rb') as f_in:
|
|
with open(mmdb_path, 'wb') as f_out:
|
|
shutil.copyfileobj(f_in, f_out, length=16 * 1024 * 1024)
|
|
print(f'Done: {mmdb_path.stat().st_size:,} bytes')
|
|
return mmdb_path
|
|
|
|
def main():
|
|
for name in ['GeoLite2-City.mmdb', 'GeoLite2-ASN.mmdb']:
|
|
gz_path = DATA_DIR / (name + '.gz')
|
|
if not gz_path.exists():
|
|
reassemble(name)
|
|
decompress(name)
|
|
print('GeoIP databases ready.')
|
|
|
|
if __name__ == '__main__':
|
|
main()
|